Regulations · Regulation (EU) 2024/1689
EU AI Act, on every AI call.
Every AI system with its risk, owner and review date, every call by a high-risk system on the record, and human oversight you can show.
In brief
Vidai applies EU AI Act to every AI call your organisation makes, inside your own network. Every AI system with a risk level, an owner and a review date; every call by a high-risk system recorded, and always in the auditor's sample; prompt injection stopped before it reaches a model. Vidai gives you the evidence; your auditor makes the call on compliance.
What Vidai enforces or records
Following EU AI Act.
- Every AI system with a risk level, an owner and a review date
- Every call by a high-risk system recorded, and always in the auditor's sample
- Prompt injection stopped before it reaches a model
- Human oversight on record: approvals, overrides and Stop AI traffic
- A serious-incident register (Enterprise)
Retention it sets
Call records and tamper-evident records kept for at least 180 days.
Coverage
Each control, shown as it really is.
| Enforced | Every key in scope has it in force. |
|---|---|
| Observed only | It's watched and logged, but some keys only observe. |
| Partly | Some keys have it and some don't, with how many. |
| Nothing in place | Nothing in scope provides it yet. |
| Your organisation's process | Vidai doesn't cover it. Listed so you see what isn't covered: always last, never hidden. |
Which controls a setting covers is our reading, not legal advice; the official text decides, and your auditor makes the call on compliance. Coverage counts AI calls sent through Vidai, not your other systems.
Where it applies
Questions
Straight answers.
- Does Vidai make us compliant with EU AI Act?
- No tool can, and Vidai never claims it. Vidai applies EU AI Act to every AI call and produces the evidence; your auditor or supervisor makes the call. A 20-minute demo shows it on infrastructure like yours, and Community edition is free to try.
- What does Vidai enforce or record for EU AI Act?
- Every AI system with a risk level, an owner and a review date. Every call by a high-risk system recorded, and always in the auditor's sample. Prompt injection stopped before it reaches a model. Human oversight on record: approvals, overrides and Stop AI traffic. A serious-incident register (Enterprise).
- How long are records kept under EU AI Act?
- Call records and tamper-evident records kept for at least 180 days. These are starting values your compliance team confirms, and a period is never shortened behind your back.
- Are EU AI Act's controls mapped in Vidai?
- Mapped control by control. The mapping is in beta: our reading of the text, which decides.
How we start together
Start with one team, on your own AI traffic.
- 01
Pick one app or team
Point it at the Control room. One configuration change; nothing else changes.
- 02
Watch for two weeks
Every call priced, checked and recorded. Rules start in observe mode.
- 03
See the evidence
What it cost, what it caught and what it would have stopped, from your own traffic.